In order to evaluate mechanisms of normal tissue injury, adequate in vivo models must be developed. Cell culture does not provide the complex environment that is found in tissues thought to be responsible for the initiation of radiation injury. In addition, experiments assessing late toxicity often require 6 months to determine if the expected injury has occurred. The delivery of radiation with these experiments must be precisely localized to the tissue of interest to prevent possible peripheral effects to confound results. Our laboratory has established an animal program for evaluation of late normal tissue toxicity through initiation of a number of animal protocols designed to develop and further study acute and late toxicity in the esophagus, lung, intestine, and muscle. This has involved the development of specialized radiation treatment immobilizers and shields to deliver the intended dose accurately. Animals have been treated with doses of radiation that we found could reproducibly result in toxicity and samples have been collected for additional high-throughput and hypothesis-driven work to determine the temporal activation of known and yet undescribed pathways in the process of radiation toxicity. In addition, a clinical trial has been initiated, NCI 07-C-0111, that will allow the collection of various biospecimens in patients receiving radiotherapy for gastrointestinal malignancies. A number of candidate biomarkers of radiation toxicity will be tested in the context of this clinical trial.